Abstract

Movement learning is complex, involving multiple structures including cortex, cerebellum and the basal ganglia. In idiopathic Parkinson's disease (PD) there is initial loss of dopaminergic innervation to the caudal putamen, which governs habitual movement. With disease progression, however, this spreads to involve more anterior regions of the basal ganglia involved in goal-directed behaviour.

Given the loss of phasic dopamine signaling in these areas we expect motor learning to be impaired in PD. The goal of this study is to investigate movement learning in PD and healthy controls using a computer-based action acquisition task.

A cohort of PD patients and age-matched healthy controls were asked to repeatedly manipulate a joystick in order to move an unseen cursor on a computer screen, initially to a seen target (task 1) and then to an unseen target (task 2). By examining how these movements are refined in order to locate the target from task 1 to task 2 (which requires action selection by the basal ganglia) we quantify the rate at which movements are learned.

The results show differences in parameters associated with the execution of the action acquisition task for the PD cohort when compared to healthy controls, supporting impaired motor learning in PD.